impact of parental migration on children’s …
TRANSCRIPT
Sri Lankan Journal of Business Economics, 2021 10 (1)
1
IMPACT OF PARENTAL MIGRATION ON CHILDREN’S
EDUCATIONAL PERFORMANCE: EVIDENCE FROM SRI LANKA
Dunusinghe P.1
Abstract
International labour migration has been one of the key sources of
employment generation and foreign exchange earnings for Sri Lanka
since early 1980s. A sizable share of the total migrants fall into the
25-44 age-group and their migration may have implications on
education and protection of children left behind. This study aimed at
examining the impact of parental migration on children left behind.
By employing a mixed method, this study collected quantitative data
by administering a questionnaire for a sample of randomly selected
households and key informant interviews and in-depth interviews
were conducted in collecting qualitative data. Both descriptive and
regression analyses were employed in data analysis and results
obtained from the above techniques were further enriched by in-
cooperating insights from key informant interviews and in-depth
interviews. Insights from the descriptive analysis, key informant
interviews, and in-depth interviews highlighted some suggestive
evidence that children in migrant households face difficulties with
their school attendance and performance compared to children in
non-migrant households. Aiming at strengthening the above
findings, this study employed a logistic regression framework in
examining the effect of migration of educational performance where
the dependent variable was self-assessment on performance at
school-level test. A number of explanatory variables were introduced
in order to control household level factors affecting educational
performance. The results clearly indicated that migration status has
no effect on educational achievement. Nevertheless, the researchers
found strong statistical evidence to suggest that educational
performance is lower among children whose mother has migrated. In
other words, probability is high that children households of mother
being a migrant to perform poorly in tests conducted at school-level.
Mothers get involved through a number of ways to improve
children’s educational achievements and their absence at home
negatively affects children’s educational performance. It is
imperative that policy makers pay urgent attention to mitigate the
above impacts.
Keywords: Labour Migration, Children Left-behind, Education,
Logit Model, Sri Lanka
1Department of Economics, University of Colombo, Sri Lanka. Email: [email protected]
Sri Lankan Journal of Business Economics, 2021 10 (1)
2
1. Introduction
Migration can be defined as movement of people from their roots to other locations,
whether it is inside or outside the country. Increasingly, migration has become a
global phenomenon, concerted with the functioning of the global capital market, with
global trade in goods and services, flow of investment, and the quality of life in
different parts of the world. Over the last few decades, the pace of international
migration has increased considerably, playing an important role in shaping and
reshaping the demographic features of countries and determining the pace of their
economic growth and development2. In Sri Lanka, during the last few decades, the
number of migrants increased rapidly due to both demand and supply factors. In 2018,
around 1.5 million Sri Lankans migrated for various purposes and one out of every
seven migrants seek employment opportunities. In Sri Lanka, the share of the females
in total labour migration has remained relatively higher than that of the male until
around 20093. Several factors such as general improvement in economic conditions,
changes in government policy on female labour migration, and higher demand for Sri
Lanka’s male workers abroad resulted in an increase of the share of males in total
out-migration4.
These imply that a sizable number of migrants are parents and they left
behind children when migrating for work. Although parental migration often brings
higher income potentially more economic resources for children's education and
health investment, lack of parental care may have adverse effects on children who are
left-behind. The net effect of parental migration depends on the relative magnitude of
the two effects and it is largely an empirical issue. In the context of developing
countries, it is imperative to identify the effect of parental migration on children’s
education, partly due to the fact that many developing countries have identified
international labour migration as one of the major avenues of reducing domestic
unemployment and enhancing the living standards of general masses and any
negative effects may lead to a vicious cycle. In particular, low human capital
formation among children left-behind could lead to another round of labour migration
thereby perpetuating the society as a whole. Moreover, factors such as demographic
transition - from extensive family structure to intensive family structure-, absence of
a public funded well-functioning child protection system, wide-spread crime
involving children, and an increasing individual-centric social formation may
increase the vulnerability towards low human capital formation among children left-
behind.
2 The United Nations (UN) estimates that the number of international migrants reached at 272
million in 2018, representing about 3 per cent of the world’s population. (United Nations
Department of Economic and Social Affairs, 2010). Between 2000 and 2018, the number of
migrants originating from middle-income countries increased faster than those from the other
income categories. 3 For instance, female migrants accounted for 73 per cent in 1995 and this share, though
declined over the years, continues to remain high (50 per cent in 2018) to this date. 4 In 2014, government of Sri Lanka makes it compulsory to submit a family background report
when a married female intend to migrate for work. Around 55 per cent of total female migrants
fall into the age category of 25-44, the childbearing age group. Similarly, around 65 per cent
of total male migrant fall into the age group of 25-44.
Sri Lankan Journal of Business Economics, 2021 10 (1)
3
Research objectives
The main objective of this study was to examine the impact of parental migration on
children’s education. Specifically, this study attempted to (a) examine how the ‘study
environment’ experienced by a child in a migrant household differ from that of one
in the non-migrant household (b) evaluate the impact of parental migration on
children’s school performance, and (c) evaluate whether the impact differs on the
basis of who migrated for work.
The section two discusses both theoretical and empirical literature on the
impact of parental migration of educational performance of children left behind while
a discussion on methodology is presented in section three. Section four presents and
discusses results derived from both descriptive and regression analyses while the last
section makes some concluding remarks.
2. Literature Review
United Nations Convention on Rights of the Child adopted in 1989 defines a 'child'
as a person below the age of 18. It was ratified by many countries including Sri Lanka,
to protect children from economic and sexual exploitation, violence, and other forms
of abuse and to enhance the rights of children to education, health care, and a decent
standard of living. These rights could be grouped into four categories namely;
survival rights, development rights, protection rights, and participation rights. The
study adopts the right based approach in examining the effect of parental migration
on children’s education, one of the pivotal rights. According to theoretical insights in
Sociology, Anthropology, and Psychology, the family (the household) plays a key
role in guaranteeing child rights that come under the above four sub-categories.
Incentives, in various forms such as instructions, rewards, care, love, kindness, and
punishments, given by family members shape the personality and the development.
The absence of parent (either one or both) may create a vacuum at household level
thereby leading to violate some social negatives. A number of sociological
perspectives (theories) guide the present study. Those theoretical models attempt to
explain theoretical underpinning related to what are children learning and how do
they learn? When do children develop emotional capabilities similar to adults? How
do children process information? How do childhood experiences affect personality
development?
The functionalist view of socialization theory, pronounces socialization as
older generations pass culture on to younger generation (Mackie, 2002). This model
assumes that a child is passive, children must be shaped to fit in, and cultural
transmission insures survival. In this respect, parents play a key role in molding
children who fit well onto existing socio-economic, political and cultural set up.
Social learning perspective, developed by Vygotsky (2004) argues children interpret,
organize, and use information from their environment. Each family provides a useful
environment at the early age of any child in social learning where they learn basics
which make the foundation for greater association with the rest of the society. This
model assumes that child development occurs in discrete stages and a child must
accomplish one stage successfully before moving onto the next. Parental migration
may affect the social learning of children left behind due to improper family
environment and as a result violation of child rights could take place.
Sri Lankan Journal of Business Economics, 2021 10 (1)
4
The interactionist perspective, developed by George and Edwards (1989)
argued that children develop identity in response to interaction with others. This
approach assumes that a child is an active agent and identity is based on children’s
perceptions of others’ perceptions of them. Absence of parents due to migration may
affect children through losing interaction with closest ones thereby affecting the
rights. The interpretive reproduction perspective, modeled by William (997), argued
that children negotiate, share and create culture with adults and each other. One of
the important assumptions of this perspective is that children influence adults in
addition to being influenced by them. Hence, parental interaction is essential in
shaping the childhood and protecting child rights. The ecological perspective,
developed by Bronfenbrenner (1986), maintains child develops competence within a
social network. On this respect, family in collaboration with schools make an
important contribution to protect child rights. The critical perspectives, developed by
Thorne (1987) and James, et. al. (1998) argued children resist domination both
actively and passively. This model assumed that children are active agents and adults
have more power than children and use it to control and dominate them. Moreover, it
is assumed that like other oppressed groups, children will resist domination. Thus,
interaction with adults (mainly parents) mold the child and maintain child rights. In
addition, psychological theory of behaviorism, emphasizes the role of reward and
punishment play in children’ learning process (Delprato & Midgley, 2002). This
theory assumes the child is passive and adults could shape the children. It could be
assumed that children in families with migrant parents may receive such reward and
punishment at sub-optimal level thereby affecting their rights. Theoretical
perspectives mentioned above shed light on the relationship between the family and
childhood where family in general and in particular parents play a critical role in
shaping the behavior and they are instruments in guaranteeing child rights including
rights to education.
The existing literature has highlighted the various channels through which
parental migration can affect the human capital development of children left-behind
(Dustman and Glitz, 2011). On the one hand, parents increase their earnings through
migration and remittance can ease the household budget constraint and thereby
increase household spending on education and reduce child labor. This theoretical
prediction has been empirically supported by studies on the effect of remittances on
children left-behind in El Salvador, Guatemala, Mexico and Philippine (Cox Edward
& Ureta, 2003; Adams & Cuecuecha, 2013; Alcaraz, Chiquiar, & Salcedo, 2012;
Yang, 2008). On the other hand, parental migration inherently leads to parental
absence from home, which can have negative effects on children left-behind through
channels such as the loss of local earnings, the lack of parenting inputs for educational
activities, and the psychological cost associated with family separation. Moreover,
parental migration also increases the migration prospects of children and can induce
more or less educational investment in children depending on the difference in the
rates of return to human capital between the migration destination and the place of
origin.
Recently, there is a growing empirical literature that examines the effects of
parental migration on educational performance of children left-behind, focusing
mainly on the dimensions of time allocations and schooling attainment (Antman,
Sri Lankan Journal of Business Economics, 2021 10 (1)
5
2012; Zhao, et. al., 2014; Lu, 2014; Nguyen, 2016; Li, et. al., 2017; Bai, et. al, 2018;
Liang and Sun, 2020; Luo, 2020) . Antman (2012), for example, found that a Mexican
father’s migration to the U.S. decreased study hours and increased work hours for
children left-behind. By examining the impact of parental migration on left-behind
children’s education across migration streams (internal vs. international), Lu (2014)
found that children left behind by international migrant parent are worse off in
educational attainment than those living with both parents. Moreover, it was found
that the impact of internal migration also remain negative in some cases, though to a
lesser extent compared to international migration. Nguyen (2016) examined the
impact of parental migration on educational performance of left-behind children in
Ethiopia, India, Peru, and Vietnam. The authors concluded that although parental
migration helps families increase per capita consumption, it does not improve health
and cognitive ability of children. Luo (2020) also found similar results for rural
China. Accordingly, the author found that direct effect of migration as well as the
indirect effect remain negative5.
Zhao et. al., (2014) discussed the impact of within country parental migration
on children’s educational performance in rural China. The authors found that math
score obtained by 4th and 5th grade students declined by around 15-60 per cent if the
student belongs to a parent migrant household. More importantly, it was found that
children in households with mother having migrated experienced a relatively higher
negative effect compared to that of the households of father having migrated. Similar
findings were made by Li, et. al., (2017) where it was found that parental migration
has a negative and significant impact on the academic performance of Junior high
school students.
A number of previous studies have looked into the effect of parental
migration on human capital formation in the context of Sri Lanka (Hettige, 1999;
Gumbard, 2008; Jampaklay, 2006; Lakshman, Perera & Sangasumana, 2015; Perera
& Jampaklay, 2014). The authors examined the effect of parental migration on
various components of human capital formation, namely on educational performance
and/or health related aspects such as nutrition. The findings of these studies are
inconclusive. Therefore, the sign and the size of the overall effect of parental
migration on the education of children left-behind is a priori unclear and remains an
empirical question. In the context of Sri Lanka, as far as the author knows, none of
the studies had examined the educational performance of left-behind children in
comparative sense, namely comparing against a comparable group of students from
non-migrant families.
5 The direct effect refers to parental accompaniment whereas the indirect effects refers to with
study time and income.
Sri Lankan Journal of Business Economics, 2021 10 (1)
6
3. Methodology
This study collected both qualitative and quantitative primary data from a sample of
households which contained both migrant households as well as non-migrant
households6. First, taking into account the district-level migration flow and ethnic
concentration, three administrative districts were selected out of the 25 districts in the
country. Second, from each selected district, the Divisional Secretarial (DS) division
reporting the highest migrant flow in 2018 was selected for the study. Third, five
Grama Niladhari (GN) divisions were selected randomly from each selected DS
division. Finally, the household questionnaire was administered at household level in
each selected GN division. The sample consisted of 450 migrant households and 150
non-migrant households in the selected districts7. Electoral lists maintained at GN
level and the lists containing migrant household information at divisional secretariat
office were utilized in drawing households for the sample. In addition, in-depth
interviews were conducted with selected children (25 in each district) in migrant
households and key public officials such as school principals, teachers, officers
responsible for migrants at divisional secretariat office, and Divisional Secretaries).
A special attention was given to the heterogeneity that exist among left-behind
children when selecting individuals for the in-depth interviews.
A number of information areas were covered by the household survey;
namely, demography, socio-economic background, family background, school
attendance, homework engagement, children’s perceptions, attending to tuition,
school performance, and school dropouts. The study interviewed children (between
10-18 years) in both migrant and non-migrant households.
This study aimed at employing both descriptive and regression analyses in
examining the impact of parental migration on educational performance of children
left-behind. Insights of key informant interviews and in-depth interviews will be
blended along with the findings from the above analyses. Following Logit model was
estimated in investigating the impact of parental migration on educational
performance;
𝑦𝑖 = 𝑋𝛽 + 𝑍𝛾 + 𝛼𝐷 + 𝑢𝑖 (Equation 1)
In equation 1, yi denotes the dependent variable that takes 1 if a student ranks
his/her educational performance as ‘average or above average’ and otherwise zero (if
a student ranks his/her educational performance as ‘below the average’)8. Matric X
includes variables representing students’ after school engagements, family
background, and school quality while Z contains variables related to migration status.
Matrix D contains district dummy variables and ui is the usual error term. β, γ, and α
are parameter vectors.
6 A migrant household is one that has at least one or both parents abroad for work at the time
of the survey and the current period of migration started six months prior to the survey was
conducted. 7 In each selected DS division, 150 migrants and 50 non-migrant households were selected
from the five GN divisions belonging to the respective DS division. 8 During the survey, students were asked to rank their educational performance and five levels
of performance were offered to them as excellent, good, average, poor, and very poor.
Sri Lankan Journal of Business Economics, 2021 10 (1)
7
4. Results and Discussion
Characteristics of household survey
Table 1 reports data on age distribution of household members by migrant and non-
migrant households in the sample. It is interesting to note that the share of relatively
small children (less than 10 years) is higher in migrant households than that in non-
migrant households for all districts except for Nuwara Eliya. Key informant
discussions revealed that economic burden is one of the main factors that forced
household members to migrate even if they have to leave behind small children.
However, this pattern is not visible for other age-groups.
Table 1: Age distribution: migrant households vs. non migrant household
Age
group
Kurunegala Nuwara Eliya Batticaloa
Non-
Migration
HH
Migration
HH
Non-
Migration
HH
Migration
HH
Non-
Migration
HH
Migration
HH
<=10 10.0 14.3 19.5 16.4 10.8 17.8
11-20 38.9 29.5 29.5 32.1 40.4 32.8
21-30 2.8 7.6 6.3 7.5 6.3 9.8
31-40 23.2 24.0 24.7 22.6 25.6 22.9
41-50 19.9 14.7 9.5 14.6 11.7 10.4
51-60 2.8 6.4 6.8 2.5 3.1 3.9
>=60 2.4 3.5 3.7 4.2 2.2 2.4
Total 100.0 100.0 100.0 100.0 100.0 100.0
Source: Based on household survey
According to our sample, in Nuwara Eliya district, ‘mother-migrated’
households accounts for around 80 per cent of total migrant households in the district
whereas this figure is around 34 per cent and 59 per cent in Batticaloa and Kurunegala
districts respectively. In contrast, share of ‘father-migrated’ households accounts for
around half of the total migrant households in Batticaloa. The ‘both-parents’ migrated
households are relatively high both in Batticaloa and Nuwara Eliya districts.
Over 94 per cent of total migrants of our sample are in Middle East and
around 60 per cent of them could be categorized as housemaids. In most cases, these
migrants may visit the family once in two years. Majority of migrated fathers and
mothers (over 70 per cent) have obtained less than secondary level of education while
there are around 5-8 per cent of total migrated fathers and mothers who have not
schooled at all. In terms of age, average age of migrated parents is 40 years while
around 5-8 per cent of migrated parents in mid 20s. Nearly 94 per cent of total
migrants remit money to their households. Those that do not send assistance are either
migrated in recent months or migrated some years ago but do not keep regular contact,
apparently disintegrated from the households. The survey found that fewer number
of mothers as well as fathers keep less contact and do not send money or other
assistance on regular basis due to limited association with the family members9. Over
9 In some cases, migrated mothers or fathers informed the family members about financial
difficulties such as not receiving monthly wage as a reason for not remitting money to the
household members.
Sri Lankan Journal of Business Economics, 2021 10 (1)
8
90 per cent of migrated mothers/fathers directly remit money to household members
on monthly basis and the remitted money is mostly utilized for financing consumption
expenditures such as food, clothes, education, health, and investing in improving
housing conditions. Children very often keep contacting migrated parents in all
districts.
Parental migration and schooling
More than 90 per cent of interviewed children (both migrant and non-migrant
households) are currently schooling. From the total sample of three districts, around
8 per cent of school-going age children in non-migrant households and 6 per cent of
school-going age children in migrant households do not school at present. Compared
to the other two districts, a sizable number of school-going students in Batticaloa
district do not school at present (see Table 2). For instance, 15 per cent of total non-
migrant household children and 9 per cent of migrant household children do not
school at present in Batticaloa. It is reasonable to guess that ‘war’ related factors may
have some influence on this situation10.
Table 2: School attendance
Kurunegala Nuwara Eliya Batticaloa Total
NM M NM M NM M NM M
Yes 96.0 96.3 92.5 94.6 85.4 91.4 91.3 94.0
No 4.0 3.7 7.5 5.4 14.6 8.6 8.7 6.0
Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
Note: NM=non-migrant and M=migrant
Source: Based on household survey
A majority of school going children attend public schools in all three districts.
In In Nuwara Eliya district, around 13 per cent of school-going children attend semi-
government schools. As revealed during the in-depth interviews, some mother
migrants get their children admitted to semi-government schools where hostel
facilities are available so that children receive better education and protection during
their absence. Most of these schools are Christian/Catholic schools. A small
percentage of children attend private schools. In most cases, children attend schools
located in nearby village or city. There is clear evidence that in Kurunegala and
Nuwara Eliya, a higher percentage of schooling children in migrant households attend
schools located in other villages or city (see Table 3). In other words, children in
migrant households attend schools located far away compared to children in non-
migrant households in these two districts. However, in Batticaloa district, a greater
percentage of children in non-migrant households attend schools that are located in
other villages/cities. The in-depth interviews revealed that migrant households are in
10 Batticaloa district was heavily affected by the civil war that lasted around 30 years. It is
possible, that relatively low schooling may associate with war-related factors such as loss of
family members, disruption of early childhood education, and lack of interest due to prolonged
absence from schools.
Sri Lankan Journal of Business Economics, 2021 10 (1)
9
a better position to bear the additional travelling costs due to sending children to
schools away from their own village/city. If it is assumed that schools located in
remote cities provide better education, one could argue that children in migrant
households access quality education due to their parents financing additional costs.
Table 3: Place of schooling
Where
school
Kurunegala Nuwara Eliya Batticaloa Total
NM M NM M NM M NM M
Living
village/City 75.0 48.9 67.6 58.6 70.7 85.7 71.4 64.7
Other
village/City 25.0 51.1 32.4 41.4 29.3 14.3 28.6 35.3
Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
Note: NM=non-migrant and M=migrant
Source: Based on household survey
Walking is the main mode of travelling for a majority of children in both
migrant and non-migrant households in Nuwara Eliya and Batticaloa districts while
public transport becomes the main mode of travelling for school children in
Kurunegala district. However, a sizable percentage of children in migrant households
travel to school by their household owned vehicles. In Kurunegala district, around
20-30 per cent of school-going children travel to schools by private school vans.
Although in-depth interviews revealed many cases where children in migrant
households travel to school by school vans data in household survey do not show that
picture. Instead, it could be seen that around 20 per cent of total children in migrant
households travel to school by household owned vehicles, mostly by bikes and three-
wheels.
Parents migration and educational difficulties
The household survey collected information related to educational difficulties
experienced by school-going children in migrant households. In addition, questions
were posed during the in-depth interviews to both school-going children as well as to
teachers/principals regarding difficulties faced by school-going children of migrant
households. In this context, focus was mainly on children’s difficulties in carrying
out homework and preparing for examinations. The household survey based results
are reported in Table 4. Accordingly, children reported difficulties related to
education after mother/father migrated11. In Kurunegala district, around 18 per cent
of children in migrated households reported difficulties with respect to education
even when mother/father or both were in Sri Lanka (before migration) and this
number increased to 40 per cent after their father/mother or both migrated (see Table
4). Similarly, in Nuwara Eliya district, around 7 per cent of children in migrated
households reported difficulties with respect to education even when mother/father
or both were in Sri Lanka (before migration) and this number increased to 29 per cent
11 Education related to difficulties mainly refer to difficulties in carrying out home-work and
preparing for examinations.
Sri Lankan Journal of Business Economics, 2021 10 (1)
10
after their father/mother or both migrated. In Batticaloa district, around 18 per cent
of children in migrated households reported difficulties with respect to education
even when mother/father or both were in Sri Lanka (before migration) and this
number increased to 54 per cent after their father/mother or both migrated. These
results imply that absence of mother/father or both has had some impact on education
related activities. In-depth interviews revealed that children faced several difficulties
in carrying out education related activities due to absence of mother/father or both.
The main among them is the absence of support for homework and preparation for
examinations.
Table 4: Parent migration and educational difficulties at school Before parent(s) migration
Kurunegala Nuwara Eliya Batticaloa Total
Yes No Total
18.4 7.4 27.0 17.7
81.6 92.6 73.0 82.3
100.0 100.0 100.0 100.0
After parent(s) migration
Kurunegala Nuwara Eliya Batticaloa Total
Yes No Total
40.0 29.1 53.7 35.0
60.0 70.9 46.3 65.0
100.0 100.0 100.0 100.0
Source: Based on household survey
In this regard, it is useful to examine children’s perception on parents’
attitude towards their education, specifically school performance. Parents’ attitude
towards children’s education is one of the key determinants of children’s success in
education. In most cases, children may perform poorly if they feel that their parents
do not care about their performance at school. During the in-depth interviews with
children, parents, and teachers/principals, it was revealed that some parents pay
greater attention on children’s education by assisting directly and/or sending for
tuition classes. This behavior is common among migrant households where father is
the migrator. Table 5 summarizes household survey results related to how children
perceive parents’ attitude towards their education before and after the mother, father,
or both migrated. In Kurunegala district, around 1 per cent of children perceived that
their parents ‘do not care about education’ even when the parents were around them
(before migration). Now (after migration), around 4 per cent of total children in
migrant households perceived that their migrated parent as well as remaining
members (mother/father/grandparents) do not care about their education. In Nuwara
Eliya district, around 9 per cent of total children felt that their parents ‘do not care
about education’ even before his/her mother/father/both migrated (before migration).
Now (after migration), around 17 per cent of total children felt their parents ‘do not
care about education’. On the other hand, in Kurunegala, around 68 per cent of total
children responded that their parents were concerned about the educational
performance and often kept contact with teachers before parent(s) migrated.
However, the percentage of children who still (after migration) feel that their parents
are concerned about education and often keep contact with teachers dropped to 57
Sri Lankan Journal of Business Economics, 2021 10 (1)
11
indicating some children increasingly feel isolated. Hence, the main message of the
above findings is that a sizable portion of children in migrated households feel that
their educational activities were moderately affected due to absence of parents
(migrated mother/father/or both). In this regard, lack of physical support and care
may work as barrier for childhood development including learning. During the in-
depth interviews, it was revealed that absence of mothers has had greater impact on
educational performance of some children, specially children in kindergarten and less
than primary. Moreover, some children do not get cleaned well and were not made
available with home-made food due to absence of their mothers. Lack of proper
cleanliness and hygiene sometimes causes children to get sick very often.
Table 5: Children’s perception on parents’ attitude toward his/her education
Parent's attitude education (before migration)
Kurunegala
Nuwara
Eliya Batticaloa Total
They do not care at all 0.7 8.8 23.0 11.2
Only care about examination marks 8.1 20.3 17.1 15.4
Concerned about school performance but
never in contact with the teachers 21.3 27.7 15.8 21.6
Concerned about school performance
and often in contact with the teachers 68.4 39.2 38.2 47.9
Do not know 1.5 4.1 5.9 3.9
Total 100.0 100.0 100.0 100.0
Parent's attitude toward education (after migration)
Kurunegala
Nuwara
Eliya Batticaloa Total
They do not care at all 4.4 16.9 23.0 15.1
Only care about examination marks 5.9 22.3 15.8 14.9
Concerned about school performance but
never in contact with the teachers 27.2 31.1 18.4 25.5
Concerned about my school performance
and often in contact with the teachers 56.6 27.0 34.2 38.8
Do not know 5.9 2.7 8.6 5.7
Total 100.0 100.0 100.0 100.0
Source: Based on household survey
Parental migration and school dropout
School dropout is relatively prominent among children in migrant households
compared to non-migrant households in all districts. In most cases, children in
migrant households tend to leave school after GCE Ordinary Level (O/L). A number
of interesting facts relating to school dropouts and poor school attendance emerged
during in-depth interviews with children and teachers/principals. Teachers and
principals mentioned that poor school attendance is prominent among children in
migrant households. Moreover, it is reported that some children in migrant
households do not show sufficient interest in education. Due to these reasons, school
performance with respect to some children in migrant households is relatively poor
resulting in school dropout at national exams such as GCE Ordinary Level. In line
with the findings of our in-depth interviews, in the household survey, ‘being weak in
Sri Lankan Journal of Business Economics, 2021 10 (1)
12
education’ and ‘exam failures’ have been cited as the main reasons for leaving
schools by school leavers in migrant households in Kurunegala and Nuwara Eliya
districts while poverty and being weak in education are the two mostly cited reasons
for school leaving in Batticaloa district (see Table 6). Even through there is weak
evidence to suggest that mother/father or both have migrated has had an impact of
school drop-out in our household survey, in-depth interview and key informant
interviews provided overwhelming evidence to suggest that to a greater extent school
dropout among children in migrant household is largely due to the absence of
father/mother or both at the household. In-depth interviews also captured some cases
of leaving school in all three districts. Lack of attention from the father/guardian often
leads to dropping out from school. It was observed that children in migrant
households receive assistance and encouragement for education when they live in an
extended family where relatives of the migrant(s) take care of educational activities.
In those cases, children in migrant households actually benefit through a number of
ways.
Table 6: Reasons for leaving school
Reasons
Kurunegala Nuwara Eliya Batticaloa Total
NM M NM M NM M NM M
Due to long distance 0.0 0.0 0.0 0.0 0.0 7.7 0.0 3.8
Poverty (lack of
income) 0.0 0.0 0.0 0.0 42.9 38.5 25.0 19.2
Being weak in
education 0.0 20.0 0.0 50.0 42.9 23.1 25.0 30.8
Exam failures 0.0 80.0 0.0 12.5 0.0 7.7 0.0 23.1
Lack of support from
the family 0.0 0.0 0.0 12.5 14.3 7.7 8.3 7.7
Other 100.0 0.0 0.0 0.0 0.0 15.4 16.7 7.7
Not responded 0.0 0.0 100.0 25.0 0.0 0.0 25.0 7.7
Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
Note: NM=non-migrant and M=migrant
Source: Based on household survey
The evidence suggests that, among school leavers, children in migrant
households have schooled less years compared to those of non-migrant households.
Out of the total sample, around 26 per cent of total school leavers have just received
less than 10 years of education.
Effect of parental migration on educational performance
Even though school attendance related data are collected, the cross-verification with
teachers, principals and key informants indicated the records obtained from school-
going kids are less reliable. During the in-depth interviews and key informant
discussions, it was found that school attendance among some children in migrant
households is relatively poor compared to children in migrant households. According
to teachers, there are a number of factors affecting this situation. Among them, lack
of support (in terms of preparing food, clothes, etc.) and attention (for homework,
attending to parent meetings, etc.) have cited as mostly affected factors. Some fathers
Sri Lankan Journal of Business Economics, 2021 10 (1)
13
in migrant households do not attend class-room meetings and their children find it
difficult to face teachers and peers when their parent do not attend important events
such as class room meetings. As a consequence, those children tend to skip school on
days on which those events are held.
Table 7: Effect of father/mother/both migration on children’s education Kurunegala Nuwara Eliya Batticaloa Total
Negatively affected 31.6 33.1 34.9 33.3
Not affected 68.4 54.7 65.1 62.6
Not responded 0.0 12.2 0.0 4.1
Total 100.0 100.0 100.0 100.0
Source: Based on household survey
Household survey asked about school-going children’s perception about
whether the parent migration affected their educational performance (see Table 7). In
this regard, attention was paid on whether their school performance (measured in
terms of marks obtained, teachers’ appreciation etc.) was affected due to the absence
of parents. Accordingly, around two-third of children believe that their mother/father
or both migration have affected their education. This fact was further reaffirmed by
the teachers/principals and key informants. It is also interesting to note that over 60
per cent of children in migrant households maintained that the migration has had no
impact on their education. It does not necessarily mean that those students perform
better in terms of education.
Key informant discussions as well as in-depth interviews revealed that school
going children in migrant families are often responsible for attending to various house
work, partly because either the mother/father or both are absent at home. Moreover,
children in migrant households are more often engaged in leisure activities largely
due to the absence of supervision from their parents. The household survey data
revealed that a significant share of children in migrant households spent time playing
with friends, watching television, engaging in economic activities compared to
children in non-migrant households (see Table 8). Around 14 per cent of children in
migrant households engage in economic activities in Batticaloa district. Lack of
attention from parents (due to migration) may be one of the reasons for engaging in
activities not related to education. Overall, over 72 per cent of children in non-migrant
households mostly engage in homework compared to 47 per cent of total children in
migrant households. This is a clear indication that children tend to engage in activities
not directly relating to education during the post-school time period. Lack of direction
and attention of parents could be one of the reasons for pushing children in migrant
households to engage in activities not directly related to education. For instance, 21
per cent of children in migrant households watch television after-school whereas this
number is around 12 per cent in non-migrant households. Moreover, it was found,
during in-depth interviews and focus group discussions, that children in migrant
households do not listen (or obey) to their guardians, in particular, when they live
with grandparents. Similarly, some school-going children in migrant households have
to play the role of a baby sitter when they have younger sisters/brothers. In some
cases, children in migrant households engage in household activities as well. For
Sri Lankan Journal of Business Economics, 2021 10 (1)
14
instance, around 8 per cent of children in migrant households engage in house work
(such as washing clothes, collecting fuel wood, cleaning kitchen items etc.) compared
to 3 per cent of children in migrant households. Children in mother migrated and both
parent migrated households, mostly engage in house work (except education related
activities). For instance, 72 per cent of children in total mother migrated households
engage in washing clothes while this number is around 60 per cent in father and both
parents migrated households. Similarly, 39 per cent of total mother and both parents
migrated households engage in cooking whereas this number is around 28 for father
migrated households. Around 20 per cent of total children in mother migrated
households and 36 per cent of total children in both parent migrated households report
that the amount of housework has increased after mother/both parents migrated.
Children in father and both parents migrated households often receive chances to
meet friends while children in mother migrated households do not receive it. For
instance, 35 per cent of children in father migrated households and 39 per cent of both
parent migrated households meet friends more often while only 26 per cent of
children in mother migrated households receive this chance. Hence, the heavy
workload born by children in mother migrated (as well as both parents migrated)
households may have a negative impact on children’s educational performance.
Table 8: After school activities Kurunegala Nuwara Eliya Batticaloa Total
NM M NM M NM M NM M
Do homework/ reading/
revision 72.0 46.3 90.0 48.6 58.3 48.0 72.5 47.7
Play with some friends 14.0 20.6 2.5 22.3 8.3 13.2 8.7 18.6
Watch TV/Listen music 14.0 27.9 7.5 23.0 14.6 13.2 12.3 21.1
Surf internet 0.0 0.0 0.0 1.4 2.1 1.3 0.7 .9
Do housework 0.0 3.7 0.0 4.1 8.3 14.5 2.9 7.6
Engage in an economic
activity 0.0 0.0 0.0 0.7 6.3 9.2 2.2 3.4
Other 0.0 1.5 0.0 0.0 2.1 0.7 0.7 0.7
Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
Source: Based on household survey
As an indirect measure of children’s educational performance, household
questionnaire asked children to grade their educational performance. According to
their own ranking, children in migrant households tend to perform poorly in education
compared to their counterparts in non-migrant households. For instance, around 10
per cent of children feel that they perform poorly whereas only 2 per cent of children
in non-migrant households feel they perform poorly in education. It is interesting to
note that around 19 per cent of total students in non-migrant households feel that they
do their studies excellently whereas only 8 per cent of total children in migrant
households feel in the similar way.
Household questionnaire as well as in-depth interviews focused on obtaining
information related to how students in migrant households feel about their
educational performance after their parents migrated. Accordingly, around 16 per
Sri Lankan Journal of Business Economics, 2021 10 (1)
15
cent of total school-going children in migrant households feel that their school
performance is worse than before. In Nuwara Eliya district, this figure stands
somewhere around 22 per cent. At the same time, it is important to note that there are
around 12 per cent of total school-going children who feel that they actually
experienced somewhat better performance after their parents migrated. Similarly,
around two-third of children in migrant households feel that their school performance
did not change after the migration. Hence, evidence against the hypothesis that
education is affected due to parent migration is not overwhelming. It is important to
mention that, as discussed elsewhere, some children in migrant households actually
perform better than previously due to the support they receive from close relatives as
well as due to their admission to tuition classes. Parent(s)’ migration has made it
possible to attend tuition classes which were not possible due to lack of financial
capacity. It was revealed that, during in-depth interviews and key informant
discussions, children living with close relatives actually perform well in education
due to the support and the supervision of the extended family members.
School performance is poorer when the length of the migration is longer. For
instance, 18 per cent of children reported that their education performance is worse
than before in households where the mother/father or both migrated less than 3 years
while this number is 24 per cent and 26 per cent for households whether mother/father
or both parents have migrated, respectively, 3-5 years and more than 5 years. When
rounds of mother migration increases, children’s educational performance become
poorer. For instance 15 per cent of children in mother migrant households reported
that their educational performance is worse than before (when the mother is in Sri
Lanka) and this number increased to 20 per cent and 22 per cent for children whose
mothers have migrated round 2 and 3 respectively. Education level of migrant mother,
father, or both parents do not affect the educational performance of children. This
might be partly due to the fact that most migrant parents report low educational
performance.
Regression analysis
The above analysis shed some light on the impact of parental migration on
educational performance of children left-behind. In order to further strengthen the
findings, a logistic regression framework, as outlined in the methodology section,
was employed in looking at the determinants of students’ educational performance.
The survey gathered data on children’s perception on their educational performance
(achievement in test) compared to their peers. The perception was measured using
five-category Likert Scale, namely Excellent, Good, Average, Poor, and Very poor.
The dependent variable is a binary variable that is constructed using above
information. It takes 1 if the perception is average or above and zero otherwise. As
per theory and empirical findings, a number of external and internal factors affect
students’ educational performance (Nisar, et. al., 2017; Betts et. al., 2014; Betts, et.
al., 2003). The external factors include household-level factors such as level of
income, parents’ education levels and occupations, level of siblings’ education,
quality of school attended, and whether or not students receive extra educational
support such as tuition. Internal, factors such as students’ commitment and attitude
towards educational activities, school attendance, past performance, and biological
Sri Lankan Journal of Business Economics, 2021 10 (1)
16
factors affect the performance. The regression analysis make use of some
explanatory variables focusing more on family environment related aspects such as
no of siblings, parent’s education levels, after-school activities, and level of
engagement with household activities. Some of the variables such as household
income was not considered partly due to the fact that the survey was administered
among children and their capacity to divulge information on household income
remains low. Both migrant and non-migrant households are in the same vicinity and
socio-economic status are mostly similar. Improvement in income, due to migration,
could be captured in the model through the binary variable.
The estimated results are reported in the model (1) through the model (6) in
Table 9. As expected in theory and in consistence with the empirical findings, the
amount of time allocated on housework (in terms of number of tasks involved) has a
negative impact on educational achievement. The estimated coefficient of
housework engagement is negative (-0.109) and significant at the conventional level
of significance in model (6). An increase in number of housework implies that
children allocate more after-school time as well as time during holidays for non-
educational household activities. As a result, time left for studies and leisure becomes
limited and, in a competitive learning environment, those students perform poorer
compared to the students who engage less in housework. The estimated coefficient
of after school activities is positive and statistically significant in the model (1)
through the model (6). In other words, educational performance is better among
students who mostly engage in educational activities during the after-school time
period. Broadly speaking, students could utilize their after-school and holiday time
for three purposes, namely for education, leisure, and housework. Students who use
time for housework mostly have limited time for education and leisure. Those who
engage less in housework could utilize their time for education and leisure. Students
both in migrant and non-migrant households could achieve better educational
outcomes provided students use more time on educational activities.
The estimated results suggest that the household head’s education level has
no effect on children’s educational outcomes. The existing literature clearly provides
evidence, both theory and empirics, that parent education level make a sizable
influence of children’s educational achievement (Nisar, et. al., 2017; Betts et. al.,
2014; Betts, et. al., 2003). The absence of evidence may be due to few factors,
namely (a) less variation in household heads’ education levels and (b) relatively low
level of education among household heads both in migrant and non-migrant
household heads.
The estimated coefficient of migration status (migrant household=1,
otherwise=0) is not statistically significant in the model (3) through the model (6).
The migration status variable captures both positive effects, such as more financial
resources for education, as well as negative effects, such as lack of close parental
supervision and support, on migrant household children’s educational achievement
compared to children in non-migrant households.
The result suggests that the net effect is not statistically meaningful.
However, descriptive data analysis showed that children in households in which the
mother has migrated face less conducive environment for studies compared to that
of the other households. Hence, the models 4 through 6 introduced a dummy variable
Sri Lankan Journal of Business Economics, 2021 10 (1)
17
that takes 1 for households with mothers having migrated and otherwise zero. The
estimated coefficient of this dummy is negative (-0.839) and statically significant at
1 per cent level of significance in the model (3) through the model (6). Mothers
influence children’s education through a number of channels. According to Baker
and Stevenson (1986), mothers influence children’s educational achievements by
encouraging children to engage in homework and school attendance, by contacting
teachers and taking appropriate actions to improve performance, and by selecting
suitable additional educational resources, such as tuition for difficult subjects. The
absence of mothers at home, hence, could negatively affect children’s educational
performance. The findings of this research are consistent with previous literature on
the effect of mother migration of educational performance of children left behind
(Dissanayake, et. al., 2014). As a sample of school children in Gampha District, the
authors found that school-level examination marks are lower for children whose
mothers are abroad compared to other children. Perera and Rathnayaka (2013) and
Save the Children (2006) found that children with absent mothers are also more
likely to drop out of schools.
Sri Lankan Journal of Business Economics, 2021 10 (1)
18
Table 9: Impact of parental migration on children’s education Dependent variable: self-assessment on performance in tests at school (1=average or above; 0=below average)
Model (1) Model (2) Model (3) Model (4) Model (5) Model (6)
Constant 0.342** -0.204 -0.089 -0.164 -0.149 -0.097
(0.179) (0.347) (0.411) (0.417) (0.427) (0.430)
Housework engagement (number of tasks involved) -0.136*** -0.145*** -0.141*** -0.126** -0.117** -0.109**
(0.043) (0.045) (0.046) (0.047) (0.048) (0.048)
After school activities (1= engaged mostly in education
activities; 0=otherwise ) 0.708*** 0.671*** 0.646*** 0.600*** 0.612*** 0.600***
(0.178) (0.183) (0.189) (0.192) (0.195) (0.196)
Number of adult siblings - 0.168** 0.164** 0.171** 0.188** 0.174**
(0.083) (0.08)3 (0.084) (0.086) (0.086)
Education of head of household (1=secondary and above,
0=less than secondary) - 0.091 0.087 0.104 0.049 0.042
(0.076) (0.076) (0.078) (0.079) (0.088)
Migration status (1=migrant household, 0=non-migrant
household) - - -0.119 0.317 0.243 0.243
(0.230) (0.263) (0.268) (0.270)
Households with mother migrated (1= mother migrant
household, 0=otherwise) - - - -0.772*** -0.773*** -0.839***
(0.215) (0.219) (0.236)
Distance to school (in KM) - - - - 0.082** 0.085**
(0.039) (0.039)
District effects No
Log likelihood value -359.68 -341.027 -340.89 -334.30 -321.46 -322.08
No of observations 546 520 520 520 509 509
Pseudo R2 0.033 0.038 0.039 0.057 0.070 0.070
Note: Standard errors are reported in parentheses while ** and *** indicate that the estimated coefficients are statistically significant at 5 per cent
and 1 per cent respectively.
Sri Lankan Journal of Business Economics, 2021 10 (1)
19
5. Conclusion
This study aimed at examining the impact of parental migration on children left
behind. After considering the size of the migration flow and ethnic composition, a
sample of households, covering both migrant and non-migrant households, was
selected from three administrative districts in Sri Lanka. Household questionnaire,
key informant interviews and in-depth interviews were conducted in collecting
relevant data for the study. Both descriptive and regression analyses were employed
in data analysis and the results obtained from the above techniques were further
enriched by in-cooperating insights from key informant interviews and in-depth
interviews.
Insights from the descriptive analysis, key informant interviews, and in-
depth interviews highlighted some suggestive evidence that children in migrant
households face difficulties with their school attendance and performance compared
to children in non-migrant households. In particular, school drop-outs were higher
among children whose either parent or both parents were abroad. Moreover, children
in migrant households were overloaded with housework, including child caring,
thereby having limited time for studies. A sizable share of children, in migrant
households, was of the opinion that their educational performance was affected due
to parental migration. These negatives were clearer among children whose mothers
were abroad. In contrast, suggestive evidence indicated that some children in migrant
households receive extra educational support, in terms of tuition, and attend schools
located in major cities due to additional financial resources in the form for
remittances received by migrant households.
Aiming at strengthening the above findings, this study employed a logistic
regression framework in examining the effect of migration of educational
performance where the dependent variable was self-assessment on performance at
school-level tests. A number of explanatory variables were introduced in order to
control household level factors affecting educational performance. The results
clearly indicated that migration status has no effect on educational performance.
Nevertheless, it was found that there are strong statistical evidence to suggest that
educational performance is lower among children whose mother has migrated. In
other words, probability is higher that children in households with mother having
migrated to perform poorly at tests conducted at school-levels. Mothers get involved
in a number of ways to improve children’s educational achievement and their
absence at home negatively affects children’s educational performance.
References
Adams, R.H. & Cuecuecha, A. (2013). The Impact of remittances on investment and
poverty in Ghana. World Development, 50, (pp. 24-40).
Alcaraz, C., Chiquiar, D. & Salcedo, A. (2012). Remittances, schooling, and child
labour in Mexico. Journal of Development Economics, 97(1), 156-165.
Antman, F. M. (2013). The impact of migration on family left behind (ch. 16). In A.
F. Constant and K. F. Zimmermann (eds.) International Handbook on the
Economics of Migration. UK: Edward Elgar.
Sri Lankan Journal of Business Economics, 2021 10 (1)
20
Bai, Y., Zhang, L., Liu, C., Shi, Y., Mo, D. & Rozelle, S. (2018). Effect of parental
migration on the academic performance of left-behind children in North
Western China. The Journal Of Development Studies, 54 (7), 1154-1170.
Baker, D. P. & Stevenson, D.L. (1986). Mothers’ strategies for children’s school
achievement: managing the transition to high school. Sociology of Education,
59(3), 156-66.
Betts, J. R., Zau, A. C. & Rice, L.A. (2003). Determinants of student achievement:
new evidence from San Diego. San Francisco: Public Policy Institute of
California.
Brayant, J. (2008). Children and international migration. in international
organization for migration (Eds.), Situation Report on International Migration
in East and South-East Asia. (pp. 177-194). Bangkok: Regional Office for
Southeast Asia.
Bronfenbrenner, U. (1986). Ecology of the family as a context for human
development: Research perspectives. Developmental psychology, 22(6), 723.
Chang, H., Dong, X. & MacPhai, F. (2011). Labor migration and time use patterns of
the left-behind children and elderly in rural China. World Development, 39
(12), 2199-2210.
Cox, E. A. & Ureta, M. (2003). International migration, remittances, and schooling:
evidence from El Salvador. Journal of Development Economics, 72(2), 429-
461.
Delprato, D. & Midgley, B.D. (2002). Some fundamentals of B.F. skinner’s
behaviorism. American Psychologist, 47(11), 1507-1520.
Dustmann, C. & Glitz, A. (2011). Migration and education. In Handbook of the
Economics of Education, 4, (pp. 327-439).
Gamburd, M.R. (2008). Milk teeth and jet planes: kin relations in families of Sri
Lanka’s transnational domestic servants. City and Society, 20(1), 5-34.
George, H. W., & Edwards, L. A. (1989). Parental attitudes toward child rearing:
instruments, issues, and implications. Psychological Bulletin, 106(1), 45-67.
Hettige, S. T. (1999). Social impact of the migration of female workers to the Middle
East. A report submitted to the Presidential Secretariat, 18 January, 1999.
James, A., Jenks, C., & Prout, A. (1998). Theorizing childhood. Cambridge: Polity
Press.
Jampaklay, A. (2006). Parental absence and children’s school enrolment: evidence
from a longitudinal study in Kanchanaburi, Thailand. Asian Population
Studies, 1(2), 69-92.
Lakshman, R., Perera, S., & Sangasumana, R. P. (2014). The children left behind by
international migrants from Sri Lanka: victims or beneficiaries of
globalization. In G. Dona and A. Veale (Eds), Children, migration &
globalization. New York: Macmillan Palgrave.
Li, L., Wang, L. & Nie, J. (2017). Effect of parental migration on the academic
performance of left-behind middle school students in rural China. China and
World Economy, 25(2), 45-59.
Liang, Z. & Sun, F. (2020). The lasting impact of parental migration on children’s
education and health outcomes: the case of China. Demographic Research,
43(9), 217-244.
Sri Lankan Journal of Business Economics, 2021 10 (1)
21
Lu, Y. (2014). Parental migration and education of left-behind children: a comparison
of two settings. Journal of Marriage and Family, 76(5), 1082-1098.
Luo, X. (2020). How does parental out-migration affect left-behind children’s
schooling outcomes? effect sizes and mechanisms. Agricultural & Applied
Economics Association, 2020. Kansas City, Missouri.
Mackie, G. (2002). Functionalist socialization, family and character, 24, pp. 40-59.
McKenzie, D., & Hillel, R. (2006). Can migration reduce educational attainment?
evidence from Mexico. Policy Research Working Paper; No. 3952. World
Bank, Washington, DC.
Nisar, N., Mahmood, M.K. & Dogar, A.H. (2017). Determinants of students'
academic achievement at secondary school level. Bulletin of Education and
Research, 39(1), 145-158.
Nguyen, C. (2016). Does parental migration really benefit left-behind children?
comparative evidence from Ethiopia, India, Peru and Vietnam. Social
Science & Medicine, vol 153, 230-239.
Perera, S. & Jampaklay, A. (2011). International migration and children left behind:
Impacts on children’s school enrolment in Sri Lanka. Sri Lanka Journal of
Population Studies, 12-13, 81-100.
Perera, N. & Rathnayaka, M. (2013). Sri Lanka’s missing mothers. Colombo: Save
the Children Sri Lanka.
Save The Children Sri Lanka, (2006). Left behind, left out: The impact on children
and families of mother migrating for work abroad. Colombo: Save the
Children.
Thorne, B. (1987). Re-visioning women and social change: Where are the children?
Gender & Society, 1(1), 85-109.
Wong, T., Ng, E.K., Yeoh, S.A.B. & Khan, H.TA. (2003). Migration and the Asian
family in a globalizing world: a selective review. Researching Migration and
the Family, 10-33.
Vygotsky, L. S. (2004). Imagination and creativity in childhood. Journal of Russian
& East European Psychology, 42(1), 7-97.
Yang, D. (2008). International migration, remittances and household investment
evidence from Philippine migrants. Economic Journal, 118(528), 591-630.